This would be a bizarre pilot suicide. Again, if you want to crash in the water, why go to the Indian Ocean. If you want to make a statement, why not hit the Petronas Tower?
Everything is possible, of course, but a lot of conclusions seem to be being drawn from "facts" that are a lot less reliable than other facts that are more reliable. It was completely freaking obvious that nobody would have any way of knowing exactly when ACARS was turned off. Just none. They would know when its last transmission was and the time at which it should-have-but-didn't make a new one, but nothing more. So
ALL that theorizing based on the ACARS being turned off before the "good night" and then the transponder being turned off after that was just completely wrong. They both could have been turned off after the good night, which is just as plausible based on what we now know.
I think we'll find that the speculation about the plane being flown above its service ceiling will also turn out to be totally wrong. Primary radar doesn't do a stellar job of determining precise altitudes, which is one reason we have transponders. Also, the significance of exceeding the service ceiling seems to be all wrong in these theories. If the thing went to 45,000 feet with a significant load, I have to think the coffin-corner would be very narrow, and it would have a hard time staying in the air, or even getting to that altitude. (Vis Pinnacle). There's no significant benefit to going that high before depressurizing. You can get the pax very easily at 35,000, so that makes no sense. If the altitude registered was actually 45,000, that speaks to me more of a wild fight for control of the aircraft, or a mistake made by an inexperienced person trying to hand-fly the thing at night with no outside reference. Deductions and inferences from these "facts" just aren't helpful.
Much better to create a hierarchy of facts based upon their perceived reliability as they stand alone, and then see how those pieces fit an existing theory or lead towards a knew one, and wait a long time before "promoting" a stray fact to a more reliable one just because it fits better into the puzzle.
What we have seem like wayward bits of data that only should be considered if they later fit into a theory that has a lot of proven facts, like the stray primary images after last secondary radar contact with the plane. Those seemed to be a red herring at first, but now SEEM to have some validity. But in the end, they may not.
Also, it seems that looking at satellite handshake signals is a very, very new procedure in aircraft accident investigations, and the deductions that have been made in that regard may turn out, upon further reflection, to be totally, totally wrong. A wayward bit of software in the satellite communication program could be creating phantom records of handshakes, for example. As a very crude and potentially implausible example, we may find that what are thought to be "records" of handshakes may be guesstimates by the satellite of where to look for the next handshake. But when complex systems and the records they create are being analyzed for purposes for which they were not intended, noise intentionally or unintentionally introduced into the system may produce records from which seeminly-plausible deductions are actually inaccurate.
One thing is for certain: once the actual facts are known (and many will be, eventually), we will be very surprised by which pieces of data actually have significance, and how they actually fit into the overall picture, just as happened with Air France. They could see indications of stuff (bending of catering carts), but how the a/c got to that point was something that nobody ever would have guessed (bunkie pulling stick full back in total-manual mode and stalling the aircraft, and continuing in a deep stall at a time when the aircraft had recovered from the loss of airspeed data and was perfectly capable of flying itself; captain not taking his seat while problem was being worked, etc. etc.). Or Colgan Air -- weather, icing, stalled T-tail???? -- nope, experienced (by the way) Captain made completely-inappropriate control inputs when faced with an emergency, stalled the aircraft, and died.
[Edited 2014-03-17 22:17:32]